A major risk factor for breast cancer is family history of the disease. Original estimates suggested that 50 percent of the cases of familial breast cancer have BRCAI and/or BRCM mutations. Further, the majority of families with breast and ovarian cancer have mutations in BRCAI. However, the frequency of BRCA mutations is not fully understood in families with a modest cancer phenotype or where other tumors occur in combination with breast. In particular Familial breast and ovarian cancer, familial breast and colon cancer and familial breast and prostate cancer are three combinations where each is known own to share common risk factors, molecular-genetic predisposition and/or shared etiologic biological plausibility. This is an application for renewal of our existing NCI project of the same name (1, U0l CA58860-04), converted from an R01 as of January 1997. To date, our current project includes population based breast (1270) and ovarian (262) cancer probands. Pedigree data on first- and second-degree relatives and first-cousins, pathology data, clinical information, epidemiologic risk and diet data, blood and tissue specimens and laboratory results are available on all probands. There will be no further ascertainment of new probands in this proposed project. The overall goal of this project is to maintain and follow up the existing family resource, further characterize BRCA mutations in breast and ovarian cancer families and explore the associated functions of BRCA1 missense mutations. In addition, we will determine whether there is molecular genetic evidence for the aggregation of breast and colorectal cancer in a subgroup of the existing high risk breast cancer families. There is strong Familial and molecular genetic evidence of an association between breast and ovarian cancer and also a familial association between breast and colon cancer. Further, we have preliminary results that show mutations in MSH2 and MLH1 in breast cancer families where there is colon cancer in first or second degree relatives. It is important, therefore, that candidate genes (such as mismatch repair genes) relevant to the tumor spectrum in these families in addition to BRCA1 and BRCA2 be examined and possible new genetic alterations be explored. The data generated from the proposed study will have future clinical applications particularly for subjects from breast cancer families, possessing the diverse spectrum of tumors as is often observed in familial breast cancer.